code-quality
Advisor for coding agents: git and Stop hooks flag test tampering, secrets and skipped hooks, with optional Jev review findings.
Overview
Code Quality is an advisor for coding agents in Claude Code, Codex and Pi, asking whether an agent that turned tests green fixed the code or the tests. Git hooks and Stop hooks raise loud red flags for secrets in changes or pushed history, deleted, skipped or weakened tests, baselines refreshed alongside code, red tests on push, and git commands that skip hooks. It also holds one complexity bar on every changed function and reports CRAP scores, diff coverage, cycles, dead code and clones across 12 languages, from one config file with no server.
It deliberately never blocks: no hook refuses a commit, push, command or end of turn; the agent sees the evidence and decides, explaining why. Jev is an optional review step, enabled in the config with a TypeSafe or OpenRouter key, that adds findings to the report alongside the deterministic checks. The listing describes it as a deterministic quality gate; the README frames it as advice.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Yes / no | Does this change have a problem worth flagging in review?(inferred) | yes · no |
How it works
- Agent edits and commits
- Hooks run deterministic checks and print red flags first
- Optional Jev review adds findings
- Agent decides and explains
Strengths
- Catches test tampering and secrets
- One config across 12 languages
- Jev optional
Caveats
- Advisory only; nothing blocks
- Jev review details are brief in the README
Getting started
Enable Jev in the [review] section of the config with jev = true and TYPESAFE_API_KEY (or OPENROUTER_API_KEY with jev_provider = "openrouter").At a glance
- Type
- Hook
- Status
- beta
- Runtime
- git hooks + agent Stop hooks
- Has code
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
Deterministic quality gate for Claude Code, Codex and pi: Git hooks that block test tampering and secrets, with an optional Jev review that adds findings to the report.
MrJev/awesome-jev
Frequently asked questions
Does it block commits?
No, it only flags; the agent decides.
What does Jev add?
Optional review findings in the report.
What counts as a red flag?
Secrets, weakened or deleted tests, refreshed baselines, red tests on push and skipped hooks.
Related projects
Coding-agent hook that asks Jev one yes/no per project rule on every edit or turn and makes the agent fix any violation.
HookbetaTypeScriptNode quality gate that scores each diff, commit message and doc set with three Jev Score lenses and a security Noul.
CLIbetaTypeScriptSemantic lint rules as yes/no questions to Jev, for review issues like logged secrets or unbounded retries that ESLint cannot express.
ProjectbetaTypeScriptSemantic code linter that reads each method with its callers and callees and asks decision models about defects, kind, line and severity.
CLIbetaTypeScriptGitHub Action that compiles .jev.yml rules into Jev questions and fails, comments on or passes a PR based on the numbers.
ProjectbetaTypeScriptFuzzy linter hook for Claude Code and Codex where Jev checks each written file against team rules and reports violations at edit time.
HookbetaTypeScript